Cell-associated adhesion molecules as early markers of ...

0 downloads 0 Views 4MB Size Report
(CD62P)/sialyl-Lewisx (CD 15s) interactions which mediate platelet-leukocyte ... platelets bound significantly onto both neutrophils and monocytes during ...
Nephrol Dial Transplant (1996) 11: 2248-2257

Nephrology Dialysis Transplantation

Original Article

Cell-associated adhesion molecules as early markers of bioincompatibility M. P. Carreno, S. Stuard, M. Bonomini, N. Settefrati, C. Tetta, A. Albertazzi and N. Haeffner-Cavaillon Institut National de la Sante et de la Recherche Medicale (INSERM U430), Hopital Broussais, 75674 Paris Cedex, Istituto di Nefrologia, Universita G. D'Annunzio, Chieti, Bellco S.p.A. Mirandola, Modena, Italy. Abstract

Background. Transient nature of adhesive interactions occurring during cell margination is mainly dependent on expression of selectins which are shed by activated cells. This shedding in the circulation may play an important role as anti-inflammatory mediator. Haemodialysis is also associated with P-selectin (CD62P)/sialyl-Lewisx (CD 15s) interactions which mediate platelet-leukocyte coaggregation. We further investigated the mechanisms underlying leukocyte margination during haemodialysis. Methods. CD 15s, CDllb and CD61 expression on circulating leukocytes from patients dialysed on synthetic membranes (modified polyacrylonitrile (SPAN), polysulphone (PS), and polyacrylonitrile (AN69) was assessed by cytofluorometry in a prospective crossover trial. We measured plasma levels C3a/C3a desArg, soluble CD62P, and CD62E molecules obtained from patients and healthy individuals. Results. Expression of CD1 lb and CD 15s was upregulated on neutrophils from patients dialysed with SPAN and PS membranes during the dialysis session. A significant negative correlation was found between the expression of CD1 lb or CD 15s molecules and neutrophil counts as well as between CD 15s expression and monocyte counts during haemodialysis. As assessed by CD61 expression on leukocytes, we observed that platelets bound significantly onto both neutrophils and monocytes during dialysis with both membranes. A significant positive correlation was found between the expression of CDllb molecules and the percentage of CD61 + monocytes counts during SPAN and PS dialysis. We found a significant increase of soluble CD62P in plasma samples obtained from haemodialysed patients before the dialysis session as compared to the levels detected in plasma from healthy individuals. Conclusions. This study documents a major role of CD 15s, CD 11 b, CD61, CD62P molecules in the transient leukocytes activation and margination during

haemodialysis on synthetic membranes despite their low complement-activating properties. Key words: adhesion molecules; biocompatibility; haemodialysis; leukocytes.

Introduction

Cell adhesion processes are activated during haemodialysis procedure and are able to initiate the leukocyte inflammatory response that itself may play an important role in the adverse effects of haemodialysis. Peripheral leukopenia during the early phase of haemodialysis is due in part to an accumulation of leukocytes in the lung vasculature [1-3]. It has been described that during inflammation adhesion molecules are involved in cell margination, activation, and subsequent transendothelial migration [4,5]. During haemodialysis using cellulosic membranes, several studies have shown that leukopenia coincides with complement activation [9-15] and is associated with an increased expression of CDllb molecule on leukocytes [16-19]. The early events of leukocytes margination involve the interaction of selectins with their ligands and we have previously reported that the transient margination observed during the haemodialysis procedure is associated with an upregulation of CD15s, a ligand of selectins, on leukocytes [20]. We have also observed that increased expression of CD 15s on leukocytes was not dependent on anaphylatoxins generation. Haemodialysis is also associated with P-selectin (CD62P)/sialyl-Lewisx (CD 15s) interactions which mediate platelet-leukocyte coaggregation [21]. Adhesion of activated platelets to leukocytes may induce a reciprocal cell activation which could mediate cellular damage to the endothelium [22-30]. The transient nature of adhesive interactions occurring during cell margination is mainly dependent on the transient Correspondence and offprint requests to: Dr Nicole Haeffner- expression of selectins which are shed by activated cells Cavaillon, INSERM U430. Hopital Broussais. 96 rue didot. 75674. [31,32]. This shedding in the circulation may play an Paris. Cedex, France

1996 European Renal Association-European Dialysis and Transplant Association

Adhesion molecules in biocompatibility

important role as anti-inflammatory mediator [33-37]. Regulation of the inflammatory processes, induced during haemodialysis may be dependent on the ratio between the amount of CD62P expressed on activated platelets and endothelial cells and that of soluble CD62P released in the blood. The purpose of our study was to further investigate the mechanisms underlying leukocyte margination during haemodialysis. Polyacrylonitrile (AN69), a new modified polyacrylonitrile (SPAN) and polysulphone (PS) were selected in a prospective crossover trial to reduce individual's variability. We have compared the expression of CD1 lb and CD 15s molecules on circulating leukocytes, the formation of platelet-leukocyte microaggregates and the production of C3a and selectins in plasma. The results indicate that despite the low-complement activating property of all membranes used, PS and SPAN membranes, there was an increased expression of CD 15s, CD61, and CDllb on circulating leukocytes from the patients. We also observed, that whatever membrane used, all patients had high levels of circulating soluble P-selectin before starting the dialysis procedure as compared to healthy individuals. The levels were not affected during the dialysis session. The plasma levels of E-selectin remained unchanged in patients dialysed with the three types of membrane and were not statistically different from those observed in healthy individuals.

Subjects and methods Patient population Six stable end-stage renal disease patients (4 men and 2 women between 43 and 67 years old; (mean + standard deviation 51+9) dialysed for more than 9 months on a stable anticoagulation regimen, and 11 healthy volunteers, were selected for study. Except for renal failure, the patients were in good health. Positive history of first use syndrome, clinical evidence of infection at the time of the study, cardiac and vascular instability, unstabilized erythropoietin dosage and single needle dialysis were exclusion criteria. None of the patients had received any medication known to affect leukocyte, platelet, or endothelial cell function for 2 weeks prior to the study. Informed consent was obtained from each patient. Study design The study covered 6 weeks of treatment, with patients treated for 2 weeks on each dialyser. The dialyser sequence was randomized. Blood samples were collected for analysis during the fifth treatment. The three membranes tested were polyacrylonitrile (Hospal H12-10, 1.3 m2, ETO-sterilised), polysulphone (Bellco, BLS 632, 1.9 m2, ETO-sterilized), and SPAN a new modified polyacrylonitrile membrane with reduced negative charges (Bellco, PA 1830, 1.8 m2, ETO sterilized). A bicarbonate-based dialysate was used in all cases. Dialysers were rinsed single-pass with 2.0 litres of 0.9% NaCl solution containing 7500 IU heparin. Heparin was the sole anticoagulant used throughout all treatments with an intermittent infusion. The patient's individual heparin regimen

2249

was maintained during the study. Blood flow rates were between 250 and 300 ml/min. Ultrafiltration rates were among 500 and 1000 ml/h. Venous blood from haemodialysed patients and healthy individuals was collected in K.3 EDTA Vacutainers and then immediately stored at 4°C before use. Blood samples were drawn predialysis (T:0) from the afferent line of the dialyser and then subsequently from the efferent line of the dialyser at 8 (T:8) and 15 (T:15)min, and at the end of the dialysis session (T:240). Dialysates were free of LPS as assessed by LAL. Reagents and buffers Phosphate-buffered saline (PBS) was from bioMerieux (Marcy PEtoile, France). Serum Albumin Bovine (BSA) and sodium azide (NaN3) were from Sigma Chemical Co. (St Louis, MO, USA). Mouse monoclonal antibodies (mAb) anti-CD 15s, anti-CD lib, anti-CD61, anti-CD62P, FITCconjugated anti-CD15, PE-conjugated anti-CD14, and isotypic Ab controls were purchased from Becton Dickinson (Mountain View, CA, USA) and Immunotech (Marseille, France). FITC-conjugated anti-mouse IgF(ab')2 fragments was obtained from Silenus Laboratories (Australia). Recombinant C5a was from Sigma and recombinant TNFa from R&D Systems (Mineapolis, USA). Flow cy tome trie analysis Neutrophils and monocytes Whole blood (300 ul) was incubated with decomplemented human AB serum (1.5ml) for 10 min at room temperature to decrease non-specific binding, and then centrifuged at 4°C. The pellet was incubated with 1 ug of first MoAb diluted in PBS-BSA 1%-sodium azide 0.1% (staining medium) or 10 ml of conjugated mAb for 30 min at 4°C. Samples were washed twice in PBS-BSA 1%. When the first mAb was unconjugated, samples were incubated with a second-affinity isolated FITC-conjugated antimouse IgF(ab')2 fragment (50 ul of 1/100 dilution) for 30 min at 4°C. Cells were washed once in staining medium. Erythrocytes were lysed by adding 2 ml of lysing solution (FACS lysing solution, Becton Dickinson) for 5 min at room temperature. This last procedure was repeated until complete lysis of red blood cells occurred. Samples were washed once with the staining medium, once with PBS. Cells were fixed by adding 0.3 ml of cold paraformaldehyde (1%) and stored in the dark at 4°C until analysis. Stained cells were characterized by flow cytometric analysis (EPICS-Elite, Coulter Electronics). The instrument was calibrated twice weekly for fluorescence and light scatter using calibrates beads (Coulter). Fluorescence parameters were collected using 3-decade logarithmic amplification. Neutrophils and monocytes were identified in the linear forward versus linear side-scatter plot by use of subgroup-specific anti-MoAbs: CD15 and CD14 for neutrophils and monocytes respectively. Mouse isotypic antibodies were used as negative controls. Platelets CD62P is a specific marker of activated platelets [37] whereas CD61 is a marker of resting and activated platelets [38]. A total of 100 ul EDTA anticoagulated whole blood was incubated for 2 h at 4°C with 1 ml of 1% cold paraformaldehyde. The fixed cells were centrifugated at 1200g for 5 min at room temperature; the pellets were washed twice in 1 ml PBS with 0.1% azide. Fifty microlitres of fixed blood suspension

M. P. Carreno el al.

2250

were incubated with 20 ul ofmAb. After 20 min incubation at room temperature, samples were washed once in 1 ml of PBS 0.1% azide. Samples were then incubated with the second anti-mouse FITC-IgF(ab')2 fragments (50 ul of 1/100 dilution) in the dark, at room temperature, for 20 min. Samples were washed once in 1 ml of PBS 0.1% azide. Platelets were fixed by adding 1 ml of cold 1% paraformaldehyde and stored in the dark at 4°C. Stained platelets were characterized by flow cytometric analysis (EPICS-Elite, Coulter Electronics). The FACScan instrument was set to measure logarithmic FSC, logarithmic SSC. Fluorescence parameters were collected using 3-decade logarithmic amplification. Platelets were identified in the logarithmic FSC versus logarithmic SSC plot by use anti-moAb CD61: the gated platelet population was found to bind >90% the antiCD61. Mouse IgGl was used as a negative control.

measures analysis of variance was used. Student-NewmanKeuls range test was applied when the F test was significant. Additional comparisons were made between SPAN, polysulphone, and polyacrylonitrile membranes at the same time points by analysis of variance followed by t test for significant differences. Student's t test was also used to compare patients with healthy individuals (unpaired data) and values obtained at different time points during in vitro studies (paired data).

Results

Leukocytes count Figure 1 depicts the changes in neutrophil and monocyte counts that occurred in patients (n = 6) haemodialysed using SPAN, polysulphone and polyacrylonitrile Platelet-leukocyte interaction (AN69) membranes during the prospective crossover MoAb-anti CD61 was used to evaluate platelet-leukocyte trial. At 8 min after the initiation of dialysis with interaction during haemodialysis with SPAN, polysulphone, SPAN membranes, there was a significant decrease in and polyacrylonitrile (AN69) membranes. The percentage of neutrophil and monocyte counts (P